Topics in Social Algorithms, MS&E 332
Spring 2018-19, Stanford University
Instructor: Ashish Goel
Contact
Email: ashishg AT stanford DOT edu
Phone: 650 eight-hundred-fourteen 1478,
Office hours: By appointment.
SYLLABUS (SUBJECT TO CHANGE)
This class presents an overview of a focused set of topics in social
algorithms; the topics change every year. The treatment is grounded in
real-life applications, but is heavy on mathematical analysis of
algorithms. The readings will largely be published papers and
surveys. We would expect algorithmic sophistication at the level of CS
261, CME 305, or MSandE212.
The emphasis in 2018-19 is going to be on social choice, market mechanisms for societal decision making, and credit networks. These
threads are inspired by practical projects at
Stanford's Social
Algorithms Lab. While the Internet has changed many socio-economic
activities, there is no online platform for substantive discussion on
complex political issues, and no way to build consensus on these
issues; this motivates the topics on large-scale decision making. Credit networks are motivated by the recent inerest in virtual currencies.
The primary requirement is one HW, a research project, and a presentation. The curriculum will be individualized to each student, and the class will meet at 1:30pm in Huang 359.
Topics covered in 2015-16:
- Credit networks and virtual currencies
- Social Choice
- An introduction to social choice theory: impossibility theorems, voting rules
- From ordinal voting to cardinal approximations
- Participatory Budgeting: Core, iterative voting, and Knapsack voting
- Opinion dynamics -- deGroot, Hegselmann-Krause, Surprising Validators, Natural algorithms, stubborn agents, biased assimilation
- Notions of Fairness
- Open problems and projects
- Market mechanisms for opinion aggregation
- Prediction markets and market scoring rules
- Lindahl equilibria
- Markets for public decision making
- The Bayesian Truth Serum
- Sequential negotitaion
- Fair splitting